X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/1e438727014e352799b4dfb8d3dfb2f8e7a58251..04007a695c00e4b9191dcd873b6a6d7b96939279:/librarian/epub.py diff --git a/librarian/epub.py b/librarian/epub.py index 2fec331..348df0c 100644 --- a/librarian/epub.py +++ b/librarian/epub.py @@ -1,13 +1,15 @@ # -*- coding: utf-8 -*- # # This file is part of Librarian, licensed under GNU Affero GPLv3 or later. -# Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information. +# Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information. # from __future__ import with_statement +from copy import deepcopy import os import os.path import subprocess +from StringIO import StringIO from copy import deepcopy from lxml import etree import zipfile @@ -16,10 +18,11 @@ from shutil import rmtree import sys -from librarian import XMLNamespace, RDFNS, DCNS, WLNS, NCXNS, OPFNS, NoDublinCore +from librarian import XMLNamespace, RDFNS, DCNS, WLNS, NCXNS, OPFNS, XHTMLNS, NoDublinCore from librarian.dcparser import BookInfo +from librarian.cover import ImageCover -from librarian import functions +from librarian import functions, get_resource functions.reg_person_name() @@ -32,7 +35,7 @@ def inner_xml(node): """ nt = node.text if node.text is not None else '' - return ''.join([nt] + [etree.tostring(child) for child in node]) + return ''.join([nt] + [etree.tostring(child) for child in node]) def set_inner_xml(node, text): """ sets node's text and children from a string @@ -73,11 +76,6 @@ def xslt(xml, sheet): return xml.xslt(etree.parse(xsltf)) -_resdir = os.path.join(os.path.dirname(os.path.realpath(__file__)), 'epub') -def res(fname): - return os.path.join(_resdir, fname) - - def replace_characters(node): def replace_chars(text): if text is None: @@ -88,13 +86,14 @@ def replace_characters(node): .replace(",,", u"\u201E")\ .replace('"', u"\u201D")\ .replace("'", u"\u2019") - if node.tag == 'extra': + if node.tag in ('uwaga', 'extra'): + t = node.tail node.clear() - else: - node.text = replace_chars(node.text) - node.tail = replace_chars(node.tail) - for child in node: - replace_characters(child) + node.tail = t + node.text = replace_chars(node.text) + node.tail = replace_chars(node.tail) + for child in node: + replace_characters(child) def find_annotations(annotations, source, part_no): @@ -110,7 +109,7 @@ def find_annotations(annotations, source, part_no): child.clear() child.tail = tail child.text = number - if child.tag not in ('extra', 'podtytul'): + if child.tag not in ('extra', 'uwaga'): find_annotations(annotations, child, part_no) @@ -257,19 +256,20 @@ def transform_chunk(chunk_xml, chunk_no, annotations, empty=False, _empty_html_s element.set('sub', str(subnumber)) if empty: if not _empty_html_static: - _empty_html_static.append(open(res('emptyChunk.html')).read()) + _empty_html_static.append(open(get_resource('epub/emptyChunk.html')).read()) chars = set() output_html = _empty_html_static[0] else: find_annotations(annotations, chunk_xml, chunk_no) replace_by_verse(chunk_xml) - html_tree = xslt(chunk_xml, res('xsltScheme.xsl')) + html_tree = xslt(chunk_xml, get_resource('epub/xsltScheme.xsl')) chars = used_chars(html_tree.getroot()) output_html = etree.tostring(html_tree, method="html", pretty_print=True) return output_html, toc, chars -def transform(provider, slug=None, file_path=None, output_file=None, output_dir=None, make_dir=False, verbose=False, sample=None): +def transform(provider, slug=None, file_path=None, output_file=None, output_dir=None, make_dir=False, verbose=False, + sample=None, cover=None, flags=None): """ produces a EPUB file provider: a DocProvider @@ -278,6 +278,8 @@ def transform(provider, slug=None, file_path=None, output_file=None, output_dir= output_dir: path to directory to save output file to; either this or output_file must be present make_dir: writes output to //.epub instead of /.epub sample=n: generate sample e-book (with at least n paragraphs) + cover: a cover.Cover object + flags: less-advertising, images, not-wl """ def transform_file(input_xml, chunk_counter=1, first=True, sample=None): @@ -293,7 +295,7 @@ def transform(provider, slug=None, file_path=None, output_file=None, output_dir= chars = set() if first: # write book title page - html_tree = xslt(input_xml, res('xsltTitle.xsl')) + html_tree = xslt(input_xml, get_resource('epub/xsltTitle.xsl')) chars = used_chars(html_tree.getroot()) zip.writestr('OPS/title.html', etree.tostring(html_tree, method="html", pretty_print=True)) @@ -301,9 +303,9 @@ def transform(provider, slug=None, file_path=None, output_file=None, output_dir= # write title page for every parent if sample is not None and sample <= 0: chars = set() - html_string = open(res('emptyChunk.html')).read() + html_string = open(get_resource('epub/emptyChunk.html')).read() else: - html_tree = xslt(input_xml, res('xsltChunkTitle.xsl')) + html_tree = xslt(input_xml, get_resource('epub/xsltChunkTitle.xsl')) chars = used_chars(html_tree.getroot()) html_string = etree.tostring(html_tree, method="html", pretty_print=True) zip.writestr('OPS/part%d.html' % chunk_counter, html_string) @@ -358,6 +360,10 @@ def transform(provider, slug=None, file_path=None, output_file=None, output_dir= raise ValueError('either slug or file_path should be specified') input_xml = etree.parse(provider[slug]) + if flags: + for flag in flags: + input_xml.getroot().set(flag, 'yes') + metadata = input_xml.find('.//'+RDFNS('Description')) if metadata is None: raise NoDublinCore('Document has no DublinCore - which is required.') @@ -378,6 +384,10 @@ def transform(provider, slug=None, file_path=None, output_file=None, output_dir= else: output_file = open(os.path.join(output_dir, os.path.splitext(os.path.basename(file_path))[0] + '.epub'), 'w') + opf = xslt(metadata, get_resource('epub/xsltContent.xsl')) + manifest = opf.find('.//' + OPFNS('manifest')) + spine = opf.find('.//' + OPFNS('spine')) + zip = zipfile.ZipFile(output_file, 'w', zipfile.ZIP_DEFLATED) # write static elements @@ -391,12 +401,53 @@ def transform(provider, slug=None, file_path=None, output_file=None, output_dir= '' \ '') - for fname in 'style.css', 'logo_wolnelektury.png': - zip.write(res(fname), os.path.join('OPS', fname)) + zip.write(get_resource('epub/style.css'), os.path.join('OPS', 'style.css')) + if not flags or 'not-wl' not in flags: + manifest.append(etree.fromstring( + '')) + zip.write(get_resource('res/wl-logo-small.png'), os.path.join('OPS', 'logo_wolnelektury.png')) + + if cover: + cover_file = StringIO() + c = cover(book_info.author.readable(), book_info.title) + c.save(cover_file) + c_name = 'cover.%s' % c.ext() + zip.writestr(os.path.join('OPS', c_name), cover_file.getvalue()) + del cover_file + + cover_tree = etree.parse(get_resource('epub/cover.html')) + cover_tree.find('//' + XHTMLNS('img')).set('src', c_name) + zip.writestr('OPS/cover.html', etree.tostring( + cover_tree, method="html", pretty_print=True)) - opf = xslt(metadata, res('xsltContent.xsl')) - manifest = opf.find('.//' + OPFNS('manifest')) - spine = opf.find('.//' + OPFNS('spine')) + manifest.append(etree.fromstring( + '')) + manifest.append(etree.fromstring( + '' % (c_name, c.mime_type()))) + spine.insert(0, etree.fromstring('')) + opf.getroot()[0].append(etree.fromstring('')) + opf.getroot().append(etree.fromstring('')) + + if flags and 'images' in flags: + for ilustr in input_xml.findall('//ilustr'): + src = ilustr.get('src') + mime = ImageCover(src)().mime_type() + zip.write(src, os.path.join('OPS', src)) + manifest.append(etree.fromstring( + '' % (src, src, mime))) + # get it up to master + after = ilustr + while after.getparent().tag not in ['powiesc', 'opowiadanie', 'liryka_l', 'liryka_lp', 'dramat_wierszowany_l', 'dramat_wierszowany_lp', 'dramat_wspolczesny']: + after = after.getparent() + if not(after is ilustr): + moved = deepcopy(ilustr) + ilustr.tag = 'extra' + ilustr.text = None + moved.tail = None + after.addnext(moved) + else: + for ilustr in input_xml.findall('//ilustr'): + ilustr.tag = 'extra' annotations = etree.Element('annotations') @@ -420,23 +471,43 @@ def transform(provider, slug=None, file_path=None, output_file=None, output_dir= nav_map.append(etree.fromstring( 'Przypisy'\ '' % {'i': toc_counter})) + toc_counter += 1 manifest.append(etree.fromstring( '')) spine.append(etree.fromstring( '')) replace_by_verse(annotations) - html_tree = xslt(annotations, res("xsltAnnotations.xsl")) + html_tree = xslt(annotations, get_resource('epub/xsltAnnotations.xsl')) chars = chars.union(used_chars(html_tree.getroot())) zip.writestr('OPS/annotations.html', etree.tostring( html_tree, method="html", pretty_print=True)) + nav_map.append(etree.fromstring( + 'Strona redakcyjna'\ + '' % {'i': toc_counter})) + manifest.append(etree.fromstring( + '')) + spine.append(etree.fromstring( + '')) + stopka = input_xml.find('//stopka') + if stopka is not None: + stopka.tag = 'stopka_' + replace_by_verse(stopka) + html_tree = xslt(stopka, get_resource('epub/xsltScheme.xsl')) + else: + html_tree = xslt(input_xml, get_resource('epub/xsltLast.xsl')) + chars.update(used_chars(html_tree.getroot())) + zip.writestr('OPS/last.html', etree.tostring( + html_tree, method="html", pretty_print=True)) + # strip fonts tmpdir = mkdtemp('-librarian-epub') cwd = os.getcwd() os.chdir(os.path.join(os.path.dirname(os.path.realpath(__file__)), 'font-optimizer')) for fname in 'DejaVuSerif.ttf', 'DejaVuSerif-Bold.ttf', 'DejaVuSerif-Italic.ttf', 'DejaVuSerif-BoldItalic.ttf': - optimizer_call = ['perl', 'subset.pl', '--chars', ''.join(chars).encode('utf-8'), res('../fonts/' + fname), os.path.join(tmpdir, fname)] + optimizer_call = ['perl', 'subset.pl', '--chars', ''.join(chars).encode('utf-8'), + get_resource('fonts/' + fname), os.path.join(tmpdir, fname)] if verbose: print "Running font-optimizer" subprocess.check_call(optimizer_call)